INDEX

Explanations

the beginning of texts or sections

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

avía

-0.50

mars

-0.50

icksburg

-0.49

 Cannot

-0.46

 Renegade

-0.46

 indi

-0.46

excel

-0.45

☒

-0.45

 lari

-0.45

unik

-0.45

POSITIVE LOGITS

 تانيه

1.41

 الوطنيه

1.00

 outta

0.85

 للمعارف

0.85

 aint

0.84

 houſe

0.84

 gonna

0.80

Literatuur

0.80

Theres

0.80

Tikang

0.79

Activations Density 0.014%

the beginning of texts or sections

No Comments

No Known Activations

the beginning of texts or sections

No Comments

No Known Activations